COIJ FOE LIBRARY
Trinity LSF
86b
WHANGPOO
CONSERVANCY BOARD
S. M. I. Series I, ISo. 2
Hydrological Data
for the
Yangtse Estuary
Obtained
up to 1918
8^6
WHANGPOO CONSERVANCY BOARD
SHANGHAI HARBOUR INVESTIGATION
(Series I. GENERAL DATA;— No. 2)
HYDROLOGICAL DATA
FOR THE
YANGTSE ESTUARY
OBTAINED
PREVIOUS TO 1918
SHANGHAI, 1919
"•■SWOesoOSC"
Printed by the COMMERCIAL PRESS, Limited
SHANGHAI
1919
PREFATORY NOTE
The hydrological data published in the previous "Report on tiie Yanglse Estuary" of
April, 1917, have since been supplemented by the Hydromelric Department, more especially by
continuous tide readings and silt-sampling at the Kiangyin and Side Saddle Stations, and further
analyzed, and an attempt is here made to provide a synopsis of the limited data available con-
veniently formulated for river and harbour engineering purposes.
To tiiis end certain generalizations have had to be mads and in some cases even assumptions
have been required to supplement facts, as the information is very scanty on many points. The
observations on which this synopsis is based only cover short and/or intermittent periods during
the Board's Estuary investigation in 1015-1916-1917 , as recorded in the above report, ivith the
exception of the folloiving permanent tide and water observation stations: —
Woosung : from 1912 to dale (Customs readings from 1880-1912).
Kiangyin : from 1915 to date.
Side Saddle : from 1915 to date.
Only further observations on a large scale can, however, remedy deficiencies. For the time
being, it has therefore been thought worth ivhile to solidify the material collected and analyzed and
when in doubt to state the deficiency but also the probable conclusions, reserving their confirmation
for future occasion.
This synopsis ivill thus show what we know and also what we do not know about the tides
and hydrography of the Estuary, which latter is equally important for future observation.
The control of the tide readings at Side Saddle has only been possible by the kind assistance
and cooperation of the Coast Inspector of the Maritime Customs, icho has allowed us the benefit of the
use of the Customs revenue steamer regularly visiting the Islands.
The preparation of these statistics, diagrams, etc., has been done by the Board's Hydromelric
Department. Dr. H. Chatley has ably supervised the ivork and assisted in preparing and editing
Oils report.
Shanghai, May, 1919.
II. von IIEIDENSTAM,
Engineer-in-Chief.
LIST OF CONTENTS
(.\) THE TIDAL WAVE AND ITS PROPAGATION:
Origin ami Direction 1
Time 1
Magnitude (Mean, Spring, Neap, and Exceptional) ... 1
Tides on the Fairy Fiats 5
Levels (Relation of Half Tide to .Mean Level, Annual Oscillation) 6
Propagation (Speed of Wave, Diminution of Range, and Change of Form) 7
Duration of Rise and Fall 9
Accuracy of Forecasts 9
(15) CURRENTS:
Origin: Tidal and Fluvial ... ... ... ... ... ... ... ... ... ... 11
Magnitudes (Means and Maxima) 11
Distribution of Velocities Vertically and Horizontally 11
Relation of Tidal Currents to Tide (Range Ratios: Difference of Phase) 14
Effect of Run-off IS
Erosion Values at Bends and Bars 15
Duration of Ebb and Flood 18
Rotary Currents at Mouth 18
(C) DISCHARGE:
Origin 19
Magnitudes (Means, Minima and Maxima) 19
Tidal Volumes ' 21
Non-tidal ("Run-off") Discharge 22
Effect on Yellow Sea 22
(D) SILT:
Origin and Character 23
Magnitude (Means, Maxima and Minima) 23
Distribution (Vertical, Horizontal, and Longitudinal) 24
Precipitation and Velocity of Equilibrium 24
Variations, Annual and Tidal 25
Total Discharge 27
Relation of Silt Density to Mud Volume 28
Density of Water Due to Silt and Salinity 28
(E) RUN-OFF:
Origin 29
Magnitude (Means, Maxima and Minima) 29
Equivalent Run-off Velocities 31
Variation from Year to Year 33
(F) SLOPE:
General Configuration of the Yangtse Valley 34
Mean-Water Profile 34
Magnitudes (Maxima and Minima) 34
Relation Between Slope and Velocity ... ... ... ... ... ... ... ... 34
"Stage" of the River 35
HIST OF PARTES
1.— The Times of High and Low Water on the Day of Full or Change of the Moon from Wuhu
to Side Saddle 2
2.— The Mean Range for All Tides from Wuhu to Side Saddle 3
3.— The Mean Ranges throughout the Lunar Month at Woosung and Side Saddle ... ... 4
4.— Standard Tidal Levels from Kiangyin to Side Saddle 5
5. — Typical Tides and Depths at the Fairy Flats (Yangtse Bar) 6
6.— The Mean Animal Oscillation or "Change of Stage" from Wuhu to Side Saddle 7
7. — Profiles of Water Surface at One Hour Intervals, Kiangyin to Side Saddle, Spring Tide ... 8
8. — The Frequency of Occurrence of a Specified Duration of a given Water Level at the Fairy
Flats 10
9. — Current Velocities from Kiangyin to Leo Point, Spring Tides 10
10. — Current Velocities from Kiangyin to Leo Point, Neap Tides 11
11.— Actual Ohserved Current Velocities in a Vertical at Leo Point 12
12. — Theoretical Relation between Range of Tide, Depth of Water, and Maximum Current
Velocity 13
13. — Relation of Velocity to Tide at Leo Point 14
14. — Computed Float Tracks in the neighbourhood of the Mouth of the Yangtse 16
15. — Computed Float Tracks in the Yellow Sea, showing the Distant Influence of the Yangtse ... 17
16.— Total Discharges in a Tidal Period, from Kiangyin to Leo Point during Spring Tides ... 10
17.— Total Discharges in a Tidal Period, from Kiangyin to Leo Point, during Neap Tides. Based
on averages of actual observations 10
18. — Approximate Sectional Areas below Standard Spring High Water, from Kiangyin to Side
Saddle 20
19. — Approximate Sectional Areas below Lowest Low Water, from Kiangyin to Side Saddle ... 20
20.— Computed Tidal Volumes passing Various Points, from Kiangyin to Drinkwater Point on a
Spring Tide Flood 21
21. — Silt Content of Water at Various Points along the Yangtse, 1917 23
22. — The Frequency of Occurrence of Specified Silt Content at the Various Stations 24
23. — Variation of Silt Content during the year at Various Stations 26
24. — Variation of Silt Discharged Past Wuhu during the year, computed from the Silt Content
and Run-off 26
25. — Variation of Mean Stage and Computed Discharge during the year at Wuhu 29
26. — Sections of Yangtse at various points 30
27. — Variation of Mean Level, Sectional Area, Discharge, and Computed Mean Velocity, at
Kiangyin, during the year 31
28.— Variations in Maximum Height, Minimum Height, and Total Range, at Kiukiang, from
1870-1915 32
20.— Mean Level, Lowest Low Water and Profile of Bottom from Kiangyin to Side Saddle ... 35
r. i i
TANGTSE ESTUARY
r'DROLOGlCAL DATA OBTAINED PREVIOUS TO 1918
(A) THE TIDAL WAVE AND ITS PROPAGATION
Origin and Direction.
The main or forced tidal wave circulates from East to West in the
Antarctic ocean and a secondary free wave proceeds toward China in a
Southeast to Northwest direction. In the China Sea the cotidal lines, i.e.,
lines connecting points where high water occurs at the same time at full and
change of the moon, indicate a general progression in the same direction,
veering north as the mouth of the Gulf of Chihli is approached (see Plate
No. 4 in Report No. 2 on the Hydrography of the Whangpoo). The speed
decreases with the depth. The wave enters the various mouths of the
Yangtse almost simultaneously (see Plate No. 35 in the Report No. 1 on the
Yangtse Estuary).
Time.
At any given locality high water occurs at a fairly constant interval
after the moon crosses the meridian which is termed the "mean establish-
ment." A table of the values of this interval for various places on the
Yangtse is given in Report No. 1 (page 32) and shows that the interval
changes from lOh. 09m. at Side Saddle to 12h. 22m. (2 minutes before the
transit) at Woosung, 5h. 0m. at Kiangyin, and 14h. 24m. at Wuhu. At new
and full moon the interval is 25 minutes longer than the mean (see Plate
No. 1).
Magnitude.
The "reference tide" employed throughout the observations is the
Woosung tide, being the one most fully observed (see preface) and is for the
Whangpoo the master tide. Full particulars of this are given in the Report
No. 2 on the Hydrography of the Whangpoo.
From this it appears that the mean tidal range is 7.30 feet, mean
high-water level 10.20 feet, mean low-water level 2.90 feet, above Woosung
Horizontal Zero, from long averages.
The mean spring range is about 10 feet and the mean neap range
about 4 feet.
[ 2 ]
Owing to the changes in the declination of the sun and moon, there is
a mean diurnal inequality in height between successive high waters of ± 1.7
feet and between successive low waters of ± 0.45 feet.
The mean tidal range (for all tides) from Side Saddle to Wuhu is
shown on Plate No. 2. Irregularities occur near Plover Point owing to the
inequalities of phase between the component waves which enter through the
two branches.
The relative ranges at different lunar epochs are in the average closely
related to the moon's hour angle. Plate No. 3 shows the mean values of the
range at Woosung and Side Saddle for the various days of the moon. (A
gross diurnal inequality of 2.5 feet between alternate pairs of ranges may
-L
J_
_L
J_
J_
300 250 200 150 100 50
DISTANCES IN GEOGRAPHICAL MILES FROM SIDE SADDLE
Plate No. 1. — The Times of High and Low Water on the Day of Full or Change
of the Moon from Wuhu to Side Saddle
L 3 ]
300
250 200 150 IOO 50
DISTANCE IN GEOGRAPHICAL MILES
Plate No. 2. — The Mean Range for All Tides from Wuhu to Side Saddle
be expected at Side Saddle at the time of maximum lunar declination.) The
values of the range at two places are fairly closely proportionate as shown in
Plate No. 2, but weak tides in the upper river are necessarily liable to
confusion with freshet pulsations there.
For engineering purposes it has been found convenient to combine
statistically the observations on the duration or frequency principle and so
standardize the tides in the Yangtse as follows: —
A Standard Spring Tide is one whose high-water and range values
are those which are exceeded by only ten per cent* of all actual tides and
whose low-water value is exceeded by 90 per cent of all actual tides.
•Ten per cent in the average obviously covers 1| days (i.e., three tides) out of the fortnightly
period, so that the underlying assumption is that the three successive tides at the correct limar epoch are
springs. Elsewhere (see Y. R. No. 1) four successive tides have been employed.
[ 4 ]
n
10
8-
o/
o/
#
h
y
H
lll
t
/ S>
7
U
I.I
U
li.
7
7
b.
z
i /
z
//
m
(/)
u
u
o
o
z
z
<
<
It
DA
Y
F
M
D
M
IT
-II
8
30 I 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 I 2
Plate No. 3. — The Mean Ranges throughout the Lunar Month at Woosung and Side Saddle
A Standard Neap Tide is one whose high-water and range values are
those which are exceeded by 90 per cent of all actual tides and whose low-
water value is attained by only ten per cent of all actual tides.
When referred to a constant level datum these definitions are not
consistent for reaches with a large variation of "stage" owing to run-off but
they are quite suitable for the lower part of an estuary.
The standard values for the entire Estuary from Side Saddle to Kiang-
yin are shown in Plate No. 4. The values at the Bar have been obtained by
the comparison of short period observations with interpolations made from
the Side Saddle and Woosung continuous records.
r 5 j
160
140
120 ioo eo eo 4-o
DISTANCE IN GEOGRAPHICAL MILES
20
Plate No. 4. — Standard Tidal Levels from Kiangyin to Side Saddle
The mean tide at the Bar in the South Channel appears to have a
range of 8.2 feet, the standard spring range heing 12.3 feet and the standard
neap 4.8 feet, but there is apparently a tendency at times to heap up on the
Bar (see Table No. 6, page 34, Yangtse Estuary Report No. 1, which gives
ranges ten per cent greater on the Bar than at Side Saddle during June,
1916).
Depths and Levels at the Bar (Fairy Flats).
The table (No. 1) shows the approximate levels and depths a
the Fairy Flats. Plate No. 5 indicates the manner in which the depths
change during a lunar day.
[ 6 ]
15
IO
**1
10
15
o*<
*V
7
?
w.
H.Z.
co
LEVt
H. H. W.
yv.
V
^ D'URN AL
NEQUAL\JY_ 6|F
L. L.W.
L OF CREST
I U R NAL
inequ4lity OF H.VM
•7
7-A-v,
?m
OF FAIRY
V.
V.
fit
o
(0
LATB
i i i i I I i i — i — i — I — I — i — I — i — i — i — I — i — i — i — r
O I I mi v H vn vm re x XI yn xmYTgwxvixvnOTiHYTXYY XYTTiniTmn'snv'KTv
Plate No. 5. — Typical Tides and Depths at the Fairy Flats (Yangtse Bar)
Levels and the "Stage" of the River.
The "half tide" level (or mean level) of any one wave may differ
Considerably from the annual mean level, since the propagation is affected
by wind and by the stage of the river due to the annual inequality of run-off.
Plate No. 6 shows the mean range of the annual oscillation of level due to
L 7 ]
20
15
RANGE OF
MEAN ANNUAL , Q
OSCILLATION OF
HALF TIDE LEVEL
(FEET) 5
250 200 150 IOO
DISTANCE IN GEOGRAPHICAL MILES
Plate No. 6. — The Mean Annual Oscillation or "Change of Stage" from Wuhu to Side Saddle
run-off (see also Table No 9, page 43, in the Yangtse Estuary Report
No. 1).
The actual rise and fall of the half tide level during any one year may
vary considerably (say as much as ± 40%) from these mean values. The
stage does not cause any appreciable variation in the tidal range in the
Estuary (it does however affect the levels to the extent of a few feet below the
Langshan Flats), but at Kiangyin the high stage tides are about 33 per cent
smaller in range than at low stage (see Plates 46 and 47, Yangtse Estuary
Report No. 1, pages 41 and 42). At Wuhu the tidal range is almost zero at
time of high water but may amount to two feet or more at low river.
Table No.
1
STANDARD WATER
LEVEL
AT FAIRY FLATS
(Nanhui Beacon)
Level reduced
to W. 11. Z.
Rue
Depth on
Crest
Seasonal Correction
to be applied for
January July
Highest High Water
...
18.0
21.8
37.8
Standard Spring High Water
...
11.8
15.6
31.6
-0.33
+1.35
Mean High Wafer
...
9.6
13.4
29.4
-0.62
+1.24
Highest Low Water
...
8.6
12.4
28.4
Standard Neap High Water...
...
7.0
10.8
26.8
-0.10
+1.27
Mean Water Level
...
5.4
9.2
25.2
Standard Neap Low Water ...
...
3.6
7.4
23.4
-1.22
+0.39
Mean Low Water
...
0.8
4.6
20.6
-0.30
+0.81
Lowest HLjh Water
...
0.2
4.0
20.0
Standard Spring Low Water
...
-0.0
3.8
19.0
-0.28
+0.28
Lowest Low Water
...
-3.8
0.0
16.0
Propagation.
From a consideration of the time of high water and low water at
the various places, the speed of propagation of the wave has been computed
t 8 J
is
C/S
2
IS
i2
I
-
o
«
w
is
t
9
on
5
—
St
S
S
o
ft-
o
35
E
[ 9 ]
and is given in Yangtse Estuary Report No. 1, Tables 5, 6*, 7, and 8 (pages
33, 34, and 35). It appears that the mean speed from Side Saddle to Wuhu
is 18^ knots for the head of the Avave and 16 for the foot, but in the open
water between Side Saddle and Nanhui speeds of 57 knots for the head and
30 knots for the foot have been observed. The depths computed from the
speeds by the standard formula agree only moderately well with the actual
deptbs.
Proceeding up the Yangtse the range steadily decreases until at Ta
Tung, 50 miles above Wuhu, it becomes imperceptible. In the summer the
limit is at Wuhu and it has not been considered necessary to consider any
tidal action above that place. The form of a typical tidal wave as far as
Kiangyin is shown on Plate 7 and there are also numerous examples given
in Report No. 1 on the Yangtse Estuary (Plates Nos. 20-34).
The rise becomes more abrupt as the river is ascended as far as
Kiangyin.
Duration of the Rise and Fall.
At Side Saddle the periods of rise and fall are practically equal,
averaging 6h. 12m. each. At Woosung the mean time of rise is 4h. 44m.
(diminishing to 3h. 50m. at springs and increasing to 5h. 38m. at neaps).
The most rapid rise occurs at Kiangyin (mean value 4h. 12m.) and
may probably be associated with the extinction of the flood current which
occurs there.
The duration of various levels over the bar is shown in Plate 8. These
levels are reduced to Woosung Horizontal Zero and to find the corresponding
depths 19f feet must be added.
Accuracy of Forecasts.
Predictions of the tidal levels at Woosung have been made by the
U. S. Geodetic and Coast Survey and also specially for the Whangpoo
Conservancy Board for the year 1918 by the Director of the Nautical
Almanac Office (London) and it appears that in the absence of strong wind
an accuracy of ± 14 minutes in time of H.W. and ± 0.6 feet in level can be
obtained. With strong winds errors of half an hour in time of H.W. and
two or three feet in level may occur. L.W. is more irregular than H.W.
•(Footnote: Table No. 6 contains a clerical error: The distance from Kiutoau Small Beacon to
Woosung should be 8.7 nautical miles, the two velocities should be 19.33 and 9.49
knots respectively and the computed depth 30.7 feet).
[ io j
Plate No. 8. — The Frequency of Occurrence of a
Specified Duration of a Given Water Level
at the Fairy Flats. 19.75 feet must be
added to find depths
(Example of Use : How often does a level of
8 feet above W. H. Z. [ = 27.75 depth]
endure for 5 hours ? Amirer : 18^% of
all tides)
Or SPECinCD Duration Of VARIOUS MMTE* LEVELS
IO 20 30 40 50 60 70
DISTANCE IN GEOGRAPHICAL MILES
Plate No. 9. — Current Velocities from Kiangyin to Leo Point, Spring Tides. This diagram is
plotted from the averages of all the actual observations. The bifurcation marked
"Division of Channel" is arbitrarily inserted to render the Pots
Tree and Leo Point values consistent
[ 11 ]
(B) CURRENTS
Origin: Tidal and Fluvial.
There are strong currents in the lower Yangtse which are due partly
to the discharge of. the run-off and partly to the transmission of the tidal
wave. The upstream or flood current is stronger than the ehb in the lower
part of the Estuary owing to the rapidity of the rise but it lasts less time. In
the upper part of the Estuary the ebb is stronger than the flood.
Magnitudes.
The information as to the values of current velocities in the Yangtse
given in the Yangtse Estuary Report No. 1 (pages 36-52) is fairly complete.
The following table of calculated means and observed maximum values (in
feet per second) taken from our actual observation series is interesting: —
(
}bs<
?rved
Calculated
Maximum
Filament Values
M
iximnm value of Mean over
whole section during
whole discharge
Flood
Ebb
Flood Ebb
Pots Tree
...
5.31
5.28
2.79 3.54
Leo Point
0.27
6.44
2.86 3.37
Kiangjin
2.70
6.30
1.87 3.92 •
The graphs show mean values for strong and weak tides as nearly as
the observations made permit (Plates 9 and 10).
*i 8
.HO
1 1 1 r- — r r»T^3
wsB»«fcJiMK L ---rTrTr~
o zo x « » «o TO *°
OISTANCE IN GtOfflWWlCAL MILES
Plate No. 10. — Current Velocities from Kiangyin to
Leo Point, Neap Tides. This diagram
is similar to Plate No. 9
Distribution of Velocities in Section.
A careful analysis of vertical velocity curves (see Yangtse Report
No. 1, pages 43-47) in divisions of the various sections shows that the
J*
[ 12 ]
VA. 2 2J4 3
3>i
4>S
VELO
CITY
IN FEI
:t pef
t SEC(
)ND
/
/
H
-
y
u
L.
Z
X
1-
(L
Q
B
OTTO
M
10
20
30
40
50
60
70
Plate No. 11. — Actual Observed Current Velocities in a Vertical at Leo Point
velocity equivalent to the mean occurs at from 30 to 67 per cent of the depth
from the bottom, but is usually about 40 per cent (i.e., 60 per cent from the
surface), and that the mean velocity in a division of the section is from 80 to
96 per cent of the actual velocity at 20 per cent of the depth from the
surface.
The bottom velocities, which are important in connection with
scouring, average as follows in relation to the mean for the whole of a
compartment of a section :
Flood Ebb
Pots Tree
Leo Point
Kiangyin
Wuhu
Values lower than 0.4 are exceptional.
An actually observed vertical velocity curve at Leo Point is shown
in Plate 11.
The transverse distribution depends on the curvature and form of the
section. In general the maximum velocities occur toward the deepest part
of the section.
0.7
0.7
0.6
0.7
0.8
0.65
—
0,80
t 13 J
2
B
^*«
n
8 SP
3^=8
2. Br P
s> e
3 OQ _
*■*• a rt
0) tsi
M "a •■
B er
•*•** -H
°"° 5"
B -» 2
*-g"<»
n a:
B -o =
Isi
B
-350
K £
„ OS
iL*
ff. **■»
o
B -|
Si' g-
if
1 °-
5 o
a? ^
H
B S
s> c_
fa
R
| |
o-c
-.3
g
OP c
g.5
£{3
B 2_
a o
V- MAXIMUM CURRENT VELOCITY IN FEET PER SECOND
5" on
o to
er-2?
"*?
K
B B
5 e
n er
a-to
c
> —
Ik
1
u
** Ol 01 -J CD <i
>
o
N
-, B
— ,„ 1 |_
to
N
LP
f
U
o
u
Ui
D-
1
*
r o
o
I
n
* /
j
/ 1
pr
/ i
/ .?
1 /
o /
> /
/
°
pi
H
II
|N)
14
1
ll
1 9
1
1
r*
I3f
/
I
oi
to "n
/ -'
jao
1
1 1
5
ll
1*
ll
f
■n o
R
\
* /
r
H
/
01
III
j
/
01
>
< 1
PI /
^1
__ /
1
*
-J
m /
PI J
H J
05
01
j
1
m
1
(X)
1
U>
I
u»
o
1
J_
1
I 1
J
[ 14 ]
Plate No. 13. — Relation of Velocity to Tide at Leo Point
It would appear that with feehle flood velocities there is a more
uniform distribution of velocity, possibly due to more extensive turbulence.
The current turns first in the weakest part of the stream so that the distribu-
tion is anomalous until the whole stream has turned.
With regard to the lateral distribution, equal velocity curves (see
Yangtse Report No. 1, pages 47-52) have been plotted from sets of
simultaneous vertical velocity observations and show that the maximum
velocities tend to occur over the deep part of the channel but may be
considerably eccentric.
Relation of Current Velocity to Tidal Wave.
A study of the Whangpoo records has confirmed (vide Whangpoo
Report No. 2, pages 3(5-38) the relation of tidal current velocity to tide
suggested by Plates 40-44 in Yangtse Estuary Report No. 1, pages 37-39,
i.e., the total range of velocity (max. ebb + max. flood velocity) in mid-
stream at one fifth the depth, is roughly proportionate to the range of the
tidal wave and when both these are measured in foot units, they happen to
be about the same value (see Whangpoo Report No. 2, appendix 2, page
80). A graph of the theoretic relation is shown in Plate 12.
[ 15 ]
A new example of this parallelism is given in Plate No. 13 taken at
random (24th Sept., 1915) from the current records for Leo Point. The
form of the two curves, range and current velocity, may differ somewhat.
A continuous record for a whole month in the Whangpoo confirms the
general constancy of this relation.
The two phenomena keep step in a fairly regular manner and in many
cases it is possible to apply the following empirical rules: —
General Characteristics of Currents in Relation to Tide
Level in Any One Cross-section.
(1) The tidal level cycle lags after the current cycle.
(2) Maximum flood velocity occurs about one hour before high water.
(3) Slack after flood occurs from one to three hours after high water
and one or two hours before half tide.
(4) Maximum ebb velocity occurs from one to five hours before low
water.
(5) Slack after ebb occurs from one half hour to three hours after
low water and from one half to one hour before half tide.
Effect of Run-off.
By subtracting the flood discharge from the ebb and dividing the
remainder by the mean area during the period and again by the total time of
flood and ebb, an equivalent mean non-tidal or run-off velocity is obtained.
The maximum values of this are as follows: —
Pots Tree 1.00 foot per second (North Branch).
Leo Point 1.35 feet (South Branch).
Kiangyin 3.92 feet (entire river).
At Wuhu where tidal effects are very small the mean of all mean
velocities observed is 3.66 feet per second and a filament value as high as
6.42 feet per second has been recorded.
It will thus be seen that in the upper river the run-off velocities
compare in value with the tidal current velocities in the lower Estuary and
that even in the latter they must appreciabby enlarge or decrease those
velocities.
Erosion Currents.
No current observations have been made at bends or bars where
erosion is known to occur but it appears probable that velocities of less than
6 feet per second produce erosion of the alluvial soil of the Yangtse Bed and
Banks.
I 16 ]
1
s
I
■ o
£
»a
a «
o _
■3 1
T-OU
z uz
o o~
7" /
o
O
u
C
s
o
*
X
1
3
'«
a
w-
V
_>»
-=
3
(8
c
-fi
O
«
O.
a Mil
-a
3 *; e s
y
IS
3
S
-a
o
3 ■,**
i
M
*2 ■ 3
■
■
'ihh
-8
o
a.
i s «lvi
a.
Sit-I'-u
a.
* " J- J
a
i!- 2 ; 5 * 1
e
£ 5 5 «
t i ° <
u
u*S S t*
9ff, .„
1
Sis- 90
■^
o
i 1 1 " 5 "
V
z o i-B^
«
u u i ;
S * ts
C-
[ 17 ]
GEO. MIS. I
SCALE OF DISPLACEMENT
12 3 4 5
6 GEO. MIS.
. «,-a T .«^. f L AT. 33* 32' N
L0CAT,0N - [LONG. 23-25- E.
161 GEO. M N. 27' 20' E.
177 GEO. M., N 4° 30' W.
H.W.<
'FLOOD 10
76 GEO. M., N. 3 9' 30' E.
N
i
location;
fLAT. 31° 57 N.
\LONG I24°40E.
147 GEO M. N. 70' 40 E.
>H.W
Note:- Figures beside the points in the curve
indicate the hours after the time of high
water at shaweishan island — lat. 3 1° 26' n long.
IZZ'lS'C; 23 GEO. M, N. 39'30'E.
Distances and directions for the places are
reckoned from the bar at the south channel
entrance to the yangtse rlver.
Plate No. 15. — Computed Float Tracks in the Yellow S»a, showing the
Distant Influence of the Yangtse
[ 18 ]
Duration of the Currents.
In the average, floods and ebbs persist for about the same periods as
the corresponding rises and falls of the tide minus and plus certain correc-
tions for the non-tidal flow. The currents have heen observed to endure for
the following periods (mean of all discharge measurements at the stations
given) : —
Flood
Ebb
Max.
Min.
Mean
Max.
Min.
Mean
Kiangyin
4h. 42m.
nil
21). 56m. ■
whole period
61). 0m.
lOh. 32m.
Leo Point
4h. 20m.
3h. Om.
3h. 52m.
81). 40m.
6h. 34m.
71). 36m,
Pots Tree
4li. 54m.
lh. 50m.
31i. 53m.
8h. 50m.
4h. 40m.
7h. 56m.
Rotary Currents.
There is no actual slack water at the mouth of the Yangtse, at least as
far up as the Bars. Transverse currents exist during the transition from ebb
to flood, thus making a rotary motion which is approximately shown in
Plates 14 and 15. Within the channel the longitudinal motions predominate
but in the open sea outside the motion is almost cycloidal, indicating the
combination of an oscillation and a drift.
Table of reputed current values as observed by mariners is also given
(Table No. 2).
Detailed tables of velocity and direction for various positions are given
on the Customs Charts and have been used to compute the rotary motion
curves referred to above.
Table No. 2
TIDAL CURRENTS IN YANGTSE
(Velocities in feet per second: Ebbs positive)
Interval from Interval from
Place Max. Ebb Max. Flood
to L. II . to H. II.
Sha Wei Shan
2h.
0m.
3h
, 0m.
Ariadne Rocks
2h.
30m.
21).
30m.
Tungsha Light
2h.
45in.
Mi.
0m.
Woosung
3h.
0m.
lh.
45 m.
High River
Max. Ebb Max. Flood
Lower Hirer p„ , , „
Max. Ebb. Max. Flood ™ marL *
7.0 (8)
4.2 (N)
-4.2 (S)
-3.4 (N)
7.0 (8)
4.2 (N)
-4.2 (S) j
-3.4 (N) "China
5.0
-6.0
5.0
-6.0 > Sea
10.0 (8)
5.0 (N)
-5.0 (8)
-3.5 (N)
10.0 (S)
5.0 (N)
-5.0 (S) ™ ot "
-3.5 (N) )
7.0 (8)
3.5 (N)
-6.0 (8)
-3.0 (N)
6 5 (S)
3.25 (N)
-6.25 ("Yangtse
-3.25 \ Pilot"
[ 19 ]
(C) DISCHARGE
Origin.
The flow past any point in the channel arises from two distinct causes
which bear a different ratio to one another at different points. These two
causes are, of course, run-off and tide. Since the tide is always an important
factor below Wuhu all observations of discharge have been taken for tidal
periods (i.e., semi-lunar clays).
Magnitude.
The total volume discharged past any section in a tidal period ("ebb"
or "flood") depends on the amount of run-off which comes through with
the ebb, and on the tidal volume which enters (on the flood) or leaves (on
the ebb) the "tidal magazine" above the section.
From Table 11 in the Yangtse Estuary Report No. 1, page 52, the
following figures for maximum observed tidal discharge for an ebb or a flood
period (a fraction of 12-j hours) are found, in millions of cubic feet.
Kiangyin
Pots Tree (North Brancli only)
Leo Point (South Branch only)
Flood
12,060
20,260
41,620
Ebb
173,050 in 24 hours
90,020 in V21 hours
35,360
97,630
The average discharges obtained are shown in Plates 16 and 17.
GEOGRAPHICAL MILES
10 30 AO 50 60 70 SO
DISTANCE IN GEOGRAPHICAL MILES
Plate No. 16. — Total Discharges in a Tidal
Period, from Kiangyin to Leo Point,
during Spring Tides. Based on
averages of actual observations
Plate No. 17. — Total Discharges in a Tidal
Period, from Kiangyin to Leo Point,
during Neap Tides. Based on
averages of actual observations
r 20 ]
3,000,000
4,500,000
z.ooqooo £
1,500,000 -J
1,000,000 5
soqooo
TO 60 5b 40 30
DISTANCE IN GEOGRAPHICAL MILES
Plate No. 18. — Approximate Sectional Areas below Standard Spring High Water,
from Kiangyin to Side Saddle
^6 /"
1
i
■
p
>
z
1
1
j /
L/ >A/|/\j vflH£
I*
%^-
XL
2,ooopoo
1,500,000 tj
^000,000 j
500,000 J
70 60 SO 40 30
DISTANCE IN GEOGRAPHICAL MILES
Plate No. 19. — Approximate Sectional Areas below Lowest Low Water
from Kiangyin to Side Saddle
[ 21 ]
SO 84 78 72 66 60 5* 48 42 36 30 24 18 12 6 O
KIANGYIN TUNG CHOW PLOVER POINT LEO POINT WOOSUNG DRINK WATER PJ
DISTANCE. IN MILES
Plate No. 20. — Computed Tidal Volumes passing Various Points from Kiangyin to
Drinkwater Point on a Spring Tide Flood
Tidal Volumes.
The volume of water which enters into a tidal river is determined by
the range of the tide, the geometrical form of the channel between high and
lower lines, the dimensions of the same, the gradient remaining from the
previous ebb modified by the run-off, the resistance to the incoming flood,
the velocities generated by the given range at the entrance under the
peculiar local conditions and the duration of the rise and fall. Hence it is
difficult to compute and all rules for doing so are empirical.
According to Stevenson's rule for tidal estuaries,* through the South
Channel near Woosung there would be about 45.24 thousand million cubic feet.
For the whole three channels it appears probable from this rule and
also our discharge measurements that the influx over the bars has a spring
tide value not much less than 250 thousand millions of cubic feet or if the
terminal plane be taken at the end of the high-water lines (the "Rusty" —
"Drinkwater Point"— "Kiutoan" line) upwards of 200,000 millions, f
This is distributed approximately as follows: —
*Tidal influx=100,000 00. ft. per sq. ft. of low water sectional area.
f Below this plane the currents on and off the banks complicate matters and are partly in the nature
of a coastal "swell."
[ 22 ]
South Channel 70,000 millions
North Channel 80,000 „
North Branch 50,000 ,,
200,000 millions
From the summation of the discharges at Pots Tree (upper end of
North Branch) and Leo Point (reduced for the volume required to raise the
water between Mason Point and Leo Point) it appears that at the junction of
the North and South Branches the whole flood wave there is not likely to
exceed some 60 thousand millions of which 20 may come from the North
Branch. The attached graphs show the sectional areas of the channels in
the Estuary (Plates Nos. 18, 19), and the approximate tidal magazine
(Plate No. 20).
Non-tidal Discharge.
The "non-tidal discharge" or "run-off" is described below (see
Sub-section E) but it may be here remarked that it can scarcely ever exceed
3 million cubic feet per second or 135 thousand millions during a tidal period
(12$ hours) and is on a yearly average only about 45 thousand millions. If
the mean tidal volume over the Bars is 100,000 millions, the mean non-tidal
discharge is about one third the whole ebb discharge over the Bars. The
observations made at the head of the North Branch may show that the ebb
divides more unecmally than the flood so that of the volume corresponding
to the run-off less than the fraction corresponding to the areas finds its way
out through the North Branch. >
Effect on the Yellow Sea.
The non-tidal flow and the reaction from the immense tidal magazine
has various effects on the Yellow Sea. It decreases the salinity for many
miles, causes a transverse current which combined with an oceanic one
produces the rotary motion already alluded to, and it actually raises the
surface in the flood season by upwards of a foot as far out as the Saddle
Islands.
Origin.
[ 23 ]
(D) SILT
The Yangtse water is charged with finely divided particles in suspen-
sion. From the fact that alluvial soil of a practically constant character
occurs right up to Shasze (Shasi), it may be deduced that originally this
silt-formed alluvium is produced from the detritus eroded from the moun-
tainous area of the Upper Yangtse. The silt present in the water is either
brought down directly from the upper torrents or has been eroded from and
replaced by similar material in the alluvial plain. The physical character of
the silt is that of a fine loam with a high proportion of sand.
800
600
400
200
WUHU
KIANGYIN
WOOSUNG SIDE SADDLE
Plate No. 21.— Silt Content of Water at Various Points along the Yangtse, 1917
Magnitude of the Charge.
The observation of silt in the Yangtse has not been continued over a'
sufficient number of years to show conclusively what the average charge is, but
the attached graph (Plate No. 21) shows what the figures were for 1917 (a
low year). It will be observed that the silt content per unit volume of water
diminishes downstream above the Estuary presumably owing to dilution of
the run-off by the tidal volume reducing the silt content at the period of
observation (high- water slack) and increases again in the Estuary owing to
the disturbance of the bottom by tidal currents. A statistical analysis of all
our observations shows that the frequency of the silt ratio at the various
points is as shown in Plate 22. The horizontals indicate what percentage of
the whole observations lie within 25 parts per million of a specified number.
[ 24 j
z
o
Z -J
&s
z a
O uj
O 0.
32
2
i zoo
IOOoJ \ — MOUTH OF YANGTSE (SUR.)
I .— MOUTH OF YANGTSE (20FT)
\— KIANGYIN (SURFACE)
4--WUHU (3 FEET)
\
V
\\\
v V
2 800
400
200
600 V
O 5 10 15 20
PERCENTAGE FREQUENCY
OF A SPECIFIED SILT CONTENT
WITHIN 1 25 PARTS PER MILLION
Plate No. 22. — The Frequency of Occurrence of Specified Silt Content at tie
Various Stations. (Example of Use: — How frequently is the silt
. content at Wuhu within ± 25 of 600 parts per million?
Answer: — One per cent of all cases)
Distribution.
There is a slight increase of the silt charge with the depth. At the
mouth of the Whangpoo the charge at 20 feet depth is upwards of ten per
cent higher than that at the surface.
From a prolonged observation at Pheasant Point in the Whangpoo it
appears that the ratio of the silt charge near the bottom to that at the surface
is about 1.66 but varies from below unity upwards according to the velocity.
No appreciable variation across a section has yet been recorded.
Longitudinally the silt charge is related in some manner to the
velocity but except for the changes noted in the previous paragraph, no data
have been obtained.
Precipitation and Silt Equilibrium.
Most of the silt consists of very fine particles which settle very slowly,
i.e., less than 1 cm. per second, the diameter being less than one tenth of a
millimeter.
The precipitation from unit volume in unit time varies approximately
as somewhat more than the square of the silt content.
[ 25 ]
The individual grains have a specific gravity of about 2.7. The dry
powder in bulk has a specific gravity of 1.25 and as wet mud 1.75. The silt
content in parts per million by weight in the Whangpoo from samples taken
at high-water slack at 20 feet depth is very approximately about 15 v 2
where v is the maximum filament velocity of the preceding current in feet
per second.
The maximum rapidity of silting on a horizontal surface in a cul-de-sac
at the mouth of the Whangpoo is about 3 feet per annum which appears to be
in excess of the silt in suspension, but may be partly due to the creep of a
thin layer of bottom mud with the more rapid flood current.
The velocity of equilibrium for a stream in alluvial soil is generally
supposed to be that which stirs up the bottom by vertical eddies generated
from the rugosities of the bed (the said eddies expanding as they rise and
descending with a much reduced vortical spin) to an extent which imparts
just as much silt to the water as is precipitated by gravity in the same time.
According to Indian canal experience this velocity varies as the two-thirds
power of the depth. This is partly supported by Yangtse, Whangpoo, and
creek experience.
The speed of precipitation is modified by bacterial culture, by addition
of alum or dilute acids, and by admixture with salt water.
Variations.
In the Estuary there are the following periodic changes in the silt
charge : —
(1) Six-hour period: — This is due to the periodic tidal current
velocity. The fine silt settles very slowly (according to experiments the rate
of precipitation varies as about the cube of the silt content) and the majority
of the silt descends less than 1 cm. per second, so that during the period of
falling velocity the actual change in silt content is small unless the previous
maximum velocity has been high and has produced a rich charge. The
large number of analyses required has prevented obtaining any very
exact relation of silt content to hourly tidal phase for the Yangtse but it
appears certain that during neap tides the silt content is almost constant
(i.e., the charge being small the rate of precipitation is very small and there
is no appreciable loss during the time when the velocity is less than that
corresponding to the silt content).
800
X
o
600
>
m
z
o
d 400
XL
Li
0.
200
1
[ 26 ]
JAN. FEB. MAR. APR. MAY JUNE JULY AUa SEPT. OCT.
Plate No. 23. — Variation of Silt Content during the year at Various
NOV. DEC.
Stations
i\
COMPU
TED |S.
ILLION
> 9
/ANNUAL 1
/ ! / \
1EAN S
LT disc
«R6E
0.
/ /
>
\
u) 40 ° H
B
2
/^ANNUAL Mi
:an silt content cur
VE ^
b
^
w o-
35
30
25
20
z
o
I-
o
z
o
-i
JAN FEB. MAR. APR. MAY JUNE JULY AUG. SEPT OCT. NOV. DEC.
Plate No. 24. — Variation of Silt Discharged past Wuhu during the year, computed
from the Silt Content and Run-off
[ 27 ]
(2) Fortnightly period: — The succession of strong currents which
occurs with spring tides raise the silt content considerably and it appears that
the ratio of the silt content on the 2d, 3d, 4th, 16th, 17th, and 18th days
of the moon to that on the 9th, 10th, 11th, 23d, 24th, and 25th is about 4.
If the three maximum and minimum daily records per half-moon are
compared, the ratio is upwards of 6. This difference is undoubtedly due to
the following causes:
(a) Maximum and minimum tidal currents do not always occur in
true tidal phase.
(b) The period of quadratures does not lie exactly midway between
syzygies.
(c) Diurnal inequality may cause acceleration or retardation, so
shifting maximum and minimum velocities away from springs and neaps, and
(d) Resultant velocities due to combined currents at any particular
point may vary according to the speed of propagation of the competent
currents.
(3) Semi-annual period : — An unexpected result is that in the lower
parts of the Estuary there are two annual maxima and minima (see Plate
No. 23) as follows:
February maximum about twice mean value.
June minimum about two-thirds mean value.
August maximum about one and a half mean value.
November minimum about half mean value.
The first maximum appears to be due to the scour which occurs at
very low river. The first minimum is a probable result of the dilution
caused by early rains and high tides. The second maximum is due to the
high stage of the river with strong currents bringing down Szechuan and
other silt. The second minimum is the result of the general decrease of the
upstream silt supply.
Total Discharge.
Volume of silt: — The annual volume of silt brought down past Wuhu
is, according to Yangtse Estuary Report No. 1 (page 65), 11,000 million cubic
feet, corresponding to an average of 16^ tons per second. A new estimate
based on the 1917 analysis and Jong period mean levels (see Plate 24) gives
an average of about 15 tons per second which at 20 cubic feet per ton comes
to almost 10,000 million cubic feet per annum. It should be clearly
[ 28 ]
understood that the above figures only apply to the freely suspended silt,
and that the amounts of silt moving along the bottom have not been taken
into account.
Relation of Silt to Mud Volume.
As the silt has a specific gravity of 2.7 the ratio by weight must be
divided by this figure to obtain the ratio by volume. In the form of mud
(specific gravity 1.75) about 40 per cent by volume is silt, the remainder
being water.
Density of Water.
The density of the Yangtse water reduced to 4° Cent, decreases from
1.0002 to 1.0010 at Wuhu to about 1.0001 at Kiangyin on account of the
decrease of silt as observed at high-water slack, 3 feet from surface, and rises
again to 1.0003 or so in the Estuary. Bej'ond the Bar the density rises
rapidly owing to the salinity and during flood tides as much as 1.025 has
been observed at Fairway Bell Buoy. This is the value in the open sea
beyond the Saddle Islands. The "Challenger" expedition found that the
density rose to a maximum of 1.026 near Japan and decreased toward the
China Coast. This decrease is undoubtedly due to the freshwater entering.
[ 29 ]
(E) RUN-OFF
Origin.
The nontidal discharge in the Yangtse originates in the precipitation
on the watershed, but a computed value has but little significance owing to
the lack of meteorological observations in the mountainous regions where the
precipitation principally occurs.
2000
1800
1600 £
IZOO o
800
u
b.
O
i/l
o
2
<
Z
5
zoo
JAN. FEB. MAR. APR MAY JUNE JULY AUG. SEPT OCT. NOV. DEC
Plate No. 25. — Variation of Mean Stage and Computed Discharge during
the year at Wuhu, 1897-1910
Magnitude.
The mean run-off at Wuhu for the period from 1897 to 1910 is
calculated to be about 1,050,000 cubic feet per second in the Yangtse Estuary-
Report No. 1, page 57. The graph, Plate 25, which has been compiled from
the mean stage curve during the year (being the average of all the years
1897-1910) and the rating curve (Plate 58, Yangtse Estuary Report No. 1,
in Folder) shows how the run-off changes through the year. The
maximum is about 3,000,000 cubic feet per second and the minimum
250,000 cubic feet per second.
[ 30 ]
t\
Z i
o
\-
u
M
h
8 Si
h
\l
[ 31 ]
©
JUNE JULY
1917
Plate No. 27. — Variation of Mean Level, Sectional Area, Discharge (Assumed Equal to that
at Wuhu) and Computed Mean Velocity, at Kiangyin, during the year 1917
About 13 per cent of the watershed lies below Wuhu but is mostly low
country with intense wet cultivation and high evaporation so that the
increase of run-off at the mouth is certainly less than ten per cent, making a
probable mean value there of about 1,100,000 and a maximum of 3,300,000
cubic feet per second.
Run-off Velocities.
The cross-sectional areas of the Estuary are shown in Plate No. 26.
The maximum and mean velocity for the whole section corresponding to the
resistance of the soil is between 3 and 5 feet per second.
At Plover Point the area at mean tide level is about 600,000 square
feet, so that the imaginary equivalent mean "run-off" velocity (i.e., the
average resultant rate of the freshwater proceeding toward the sea) is from
0.4 to 5.0 feet per second. At the Bars (outer end of Tsungming Island) the
[ 32 ]
48 -
46-
44-
42-
40-
38-
36-
34-
32-
30-
28-
26-
24-
22-
20-
18 -
16-
14-
12-
10-
8-
6 -
4-
2-
ii
l\
Ml
kX
IM|U
1
Ht
_
.,
J
\
\
1
^
f i \
'1
|
'1
\ 1
>
\
'
^
J
\l
r
\\
\
a
L
V
\
'
I
t
,; V
1 1
T
t
V-
\
i
\
\
\
i
/
t
/
I
\
\
\
•••-,
.A
\li
M
V
V
\
I
\
<
1
!
1
1
i
i
i
\ 1
W
1
■V
If
\
>
i
\
j
/•
1 ,
1,
\
t
i
H
\t
Jt,
E
' \ 1
» ' 1
J( 1 ►
1 .
;
'!
1 1
ll
i
t 1
1
:
J
II
\
j
/
J
Ml
Nl
Ml
Ji"
HE
:ig
HT
1
i
■ l
1
I
1
i
\
A
A,
1 \
j \
1
!
\
t
/
/
/
\
i
1
i
/
•
\
' \
y
V
\
'
/
\
f
1
>
\
^
/
V\
i
\
\\
K
a
K
IA
N
1
Z
'-*<.
3
\ '"*!
X
1
70
18
75
8
J
18
B5
18
30
D
^
1
E
B95
1901
V
1305
i
910
19
S
ia
Plate No. 28. — Variations in Maximum Height, Minimum Height, and Total Range,
at Kiukiang, from 1870-1915
total area of the water section is something like 3,000,000 square feet, so that
the equivalent run-off velocity there is from 0.1 to 1.0 foot per second.
Diagram of run-off velocities at Kiangyin for one year computed from
Wuhu discharges are shown in Plate 27.
It will thus be seen that the ratio of the equivalent run-off velocity to
the tidal current velocities (which amount to from 5 to 7 feet per second at
Plover Point) is from, say, 0.06 to almost unity at Plover and with a
[ 33 ]
maximum velocity of, say, 8 feet per second at the Bar, the ratio is from
0.01 to 0.13. These figures will be almost doubled in the lower parts of the
Estuary if it is considered that run-off only occurs during ebbs.
It thus appears that the "run-off" velocity is an important factor in
the maximum strength of the current, say above Woosung, but only causes
slight differences in the current below that place.
Variation from Year to Year.
The graph (Plate 27) shows- the maximum and minimum stage at
Kiukiang from 1871 to 1910 which fluctuates about 50 per cent more than
Wuhu but has the advantage of being entirely free of tidal influence. 1901
and 1911 are the two highest years.
[ 34 ]
(F) SLOPE
General Configuration of the Yangtse Valley.
Above Ichang the slopes are large and the flow is torrential but below
that place the fall is very moderate so that spate and (in the Estuary) tide
conditions produce great relative changes.
Mean Profile.
Plate No. 29 shows the mean-water profile from Wuhu to Side Saddle
and also the Highest High-Water and Lowest Low- Water Lines.
All these lines are imaginary since the tidal waves are always in
existence but they doubtless approximately correspond to the run-off. A set
of instantaneous water profiles is shown in Plate No. /l.
Magnitudes.
The maximum observed flood slope is 0.00008 near Plover Point, but
there is almost a "bore" effect with infinite flood slope at certain times over
some of the shoals. The maximum ebb slope observed is 0.00004 in the
same locality.
Relation Between Slope and Velocity.
Attempts have been made at Wuhu to correlate the velocity to the
slope and hydraulic radius on the lines of the different standard velocity
formulae but the difficulties, observational and mathematical, are very
serious. There is doubtless an acceleration correction (Whangpoo Report
No. 2, pages 67-77) and it also seems certain that "dissipation" — i.e.,
irrecoverable loss of energy due to eddying at changes of section and changes
of direction plays a large part when a considerable length of the stream is
considered. *
*The following formula fairly well expresses the conditions in a reach with equal end water sections
v = C (l-k)Ro.7(l-f)°- 5
When Co is the coefficient for an artificial regular uniform channel,
k is the dissipation factor for a natural channel.
R is hydraulic radius.
I is slope, a is dv/dt, v=mean velocity, g=gravitationaI ace.
I 35 J
Plate No. 29. — Mean Level, Lowest Low Water and Profile of Bottom of Channel
from Kiangyin to Side Saddle
The "Stage" of the River.
The "stage" at Wuhu is of course a rough indication of the slope
from there to the sea, and the discharge is related to it in a complex manner
which is described in the first Yangtse Report. Its relation to the local
slope cannot be determined without comparison with the stages at other
points above and below and the accurate levelling of the necessary tidepoles
to a degree commensurate with that of the minute differences of level to be
observed.
Shanghai, May, 1919.
H. von HEIDEJJ8TAM,
En g ineer-in- Chief.
Photomount
Date Due
nuNa iQuiruerr bureau Cat. No, I090A
Photomount
Pamphlet
Binder
Gaylord Bros., Inc.
Makers
Syracqse, N. Y.
W.ilH 21, 1908
1840 00543 4958
Live Music
Archive
Librivox
Free Audio
Metropolitan Museum
Cleveland
Museum of Art
Internet
Arcade
Console Living Room
Open Library
American
Libraries
TV News
Understanding
9/11